registry good first issues (Python)

Map AML.T0086: Exfiltration via AI Agent Tool Invocation 4 days ago

good first issue track:patterns season-1

1

An open, vendor-neutral registry of known attacks on AI agents: a stable ID and MITRE ATLAS mapping, a harmless test, and a detection, for each one.

Python

Map AML.T0118.001: Autonomous AI Agent Communication: Direct Agent Communication 5 days ago

good first issue track:patterns season-1

guardianprotocol-ai/registry

1

An open, vendor-neutral registry of known attacks on AI agents: a stable ID and MITRE ATLAS mapping, a harmless test, and a detection, for each one.

Python

Map AML.T0010.005: AI Supply Chain Compromise: AI Agent Tool 9 days ago

good first issue track:patterns season-1

guardianprotocol-ai/registry

1

An open, vendor-neutral registry of known attacks on AI agents: a stable ID and MITRE ATLAS mapping, a harmless test, and a detection, for each one.

Python

gemini-cli hangs on GP-0003, losing most of the runs 15 days ago

help wanted track:scanner

guardianprotocol-ai/registry

1

An open, vendor-neutral registry of known attacks on AI agents: a stable ID and MITRE ATLAS mapping, a harmless test, and a detection, for each one.

Python

GP-0001 and GP-0002 are the same experiment 16 days ago

good first issue track:patterns

guardianprotocol-ai/registry

1

An open, vendor-neutral registry of known attacks on AI agents: a stable ID and MITRE ATLAS mapping, a harmless test, and a detection, for each one.

Python

Measure the runnable patterns against Claude Code 16 days ago

help wanted track:scanner

guardianprotocol-ai/registry

1

An open, vendor-neutral registry of known attacks on AI agents: a stable ID and MITRE ATLAS mapping, a harmless test, and a detection, for each one.

Python

Measure your own agent 16 days ago

help wanted track:scanner

guardianprotocol-ai/registry

1

An open, vendor-neutral registry of known attacks on AI agents: a stable ID and MITRE ATLAS mapping, a harmless test, and a detection, for each one.

Python

Measure an open-weight model through Ollama 16 days ago

help wanted track:scanner

guardianprotocol-ai/registry

1

An open, vendor-neutral registry of known attacks on AI agents: a stable ID and MITRE ATLAS mapping, a harmless test, and a detection, for each one.

Python

Measure the runnable patterns against Gemini CLI 16 days ago

help wanted track:scanner

guardianprotocol-ai/registry

1

An open, vendor-neutral registry of known attacks on AI agents: a stable ID and MITRE ATLAS mapping, a harmless test, and a detection, for each one.

Python

Measure the runnable patterns against Codex CLI 16 days ago

help wanted track:scanner

guardianprotocol-ai/registry

1

An open, vendor-neutral registry of known attacks on AI agents: a stable ID and MITRE ATLAS mapping, a harmless test, and a detection, for each one.

Python

Measure the runnable patterns against Cursor 16 days ago

help wanted track:scanner

guardianprotocol-ai/registry

1

An open, vendor-neutral registry of known attacks on AI agents: a stable ID and MITRE ATLAS mapping, a harmless test, and a detection, for each one.

Python

Map AML.T0130: AI Agent Response Biasing 16 days ago

good first issue track:patterns season-1

guardianprotocol-ai/registry

1

An open, vendor-neutral registry of known attacks on AI agents: a stable ID and MITRE ATLAS mapping, a harmless test, and a detection, for each one.

Python

Map AML.T0118.000: Autonomous AI Agent Communication: Communication via Shared Artifacts 16 days ago

good first issue track:patterns season-1

guardianprotocol-ai/registry

1

An open, vendor-neutral registry of known attacks on AI agents: a stable ID and MITRE ATLAS mapping, a harmless test, and a detection, for each one.

Python

Map AML.T0118: Autonomous AI Agent Communication 16 days ago

good first issue track:patterns season-1

guardianprotocol-ai/registry

1

An open, vendor-neutral registry of known attacks on AI agents: a stable ID and MITRE ATLAS mapping, a harmless test, and a detection, for each one.

Python

Map AML.T0110.000: AI Agent Tool Poisoning: Definition and Instructions 16 days ago

good first issue track:patterns season-1

guardianprotocol-ai/registry

1

An open, vendor-neutral registry of known attacks on AI agents: a stable ID and MITRE ATLAS mapping, a harmless test, and a detection, for each one.

Python

Map AML.T0101: Data Destruction via AI Agent Tool Invocation 16 days ago

good first issue track:patterns season-1

guardianprotocol-ai/registry

1

An open, vendor-neutral registry of known attacks on AI agents: a stable ID and MITRE ATLAS mapping, a harmless test, and a detection, for each one.

Python

Map AML.T0098: AI Agent Tool Credential Harvesting 16 days ago

good first issue track:patterns season-1

guardianprotocol-ai/registry

1

An open, vendor-neutral registry of known attacks on AI agents: a stable ID and MITRE ATLAS mapping, a harmless test, and a detection, for each one.

Python

Map AML.T0083: Credentials from AI Agent Configuration 16 days ago

good first issue track:patterns season-1

guardianprotocol-ai/registry

1

An open, vendor-neutral registry of known attacks on AI agents: a stable ID and MITRE ATLAS mapping, a harmless test, and a detection, for each one.

Python

Map AML.T0080.000: AI Agent Context Poisoning: Memory 16 days ago

good first issue track:patterns season-1

guardianprotocol-ai/registry

1

An open, vendor-neutral registry of known attacks on AI agents: a stable ID and MITRE ATLAS mapping, a harmless test, and a detection, for each one.

Python

Map AML.T0080: AI Agent Context Poisoning 16 days ago

good first issue track:patterns season-1

guardianprotocol-ai/registry

1

An open, vendor-neutral registry of known attacks on AI agents: a stable ID and MITRE ATLAS mapping, a harmless test, and a detection, for each one.

Python

Map AML.T0069.002: Discover LLM System Information: System Prompt 16 days ago

good first issue track:patterns season-1

guardianprotocol-ai/registry

1

An open, vendor-neutral registry of known attacks on AI agents: a stable ID and MITRE ATLAS mapping, a harmless test, and a detection, for each one.

Python

Map AML.T0068: LLM Prompt Obfuscation 16 days ago

good first issue track:patterns season-1

guardianprotocol-ai/registry

1

An open, vendor-neutral registry of known attacks on AI agents: a stable ID and MITRE ATLAS mapping, a harmless test, and a detection, for each one.

Python

Map AML.T0061: LLM Prompt Self-Replication 16 days ago

good first issue track:patterns season-1

guardianprotocol-ai/registry

1

An open, vendor-neutral registry of known attacks on AI agents: a stable ID and MITRE ATLAS mapping, a harmless test, and a detection, for each one.

Python

Map AML.T0057: LLM Data Leakage 16 days ago

good first issue track:patterns season-1

guardianprotocol-ai/registry

1

An open, vendor-neutral registry of known attacks on AI agents: a stable ID and MITRE ATLAS mapping, a harmless test, and a detection, for each one.

Python

Map AML.T0056: Extract LLM System Prompt 16 days ago

good first issue track:patterns season-1

guardianprotocol-ai/registry

1

An open, vendor-neutral registry of known attacks on AI agents: a stable ID and MITRE ATLAS mapping, a harmless test, and a detection, for each one.

Python

Map AML.T0053: AI Agent Tool Invocation 16 days ago

good first issue track:patterns season-1

guardianprotocol-ai/registry

1

An open, vendor-neutral registry of known attacks on AI agents: a stable ID and MITRE ATLAS mapping, a harmless test, and a detection, for each one.

Python

Map AML.T0051.001: LLM Prompt Injection: Indirect 16 days ago

good first issue track:patterns season-1

guardianprotocol-ai/registry

1

An open, vendor-neutral registry of known attacks on AI agents: a stable ID and MITRE ATLAS mapping, a harmless test, and a detection, for each one.

Python

Map AML.T0034.002: Cost Harvesting: Agentic Resource Consumption 16 days ago

good first issue track:patterns season-1

guardianprotocol-ai/registry

1

An open, vendor-neutral registry of known attacks on AI agents: a stable ID and MITRE ATLAS mapping, a harmless test, and a detection, for each one.

Python

guardianprotocol-ai/registry

Open Issues Need Help